Is one interval session a week enough?
In 315 inactive adults with central obesity, 75 min/week of 4×4 intervals as one visit (three 25-min bouts with breaks) or three both raised VO2max vs control; 3 vs 1: +1.1, p = 0.058. A, Tier 0.
Main caveat: The frequency contrast is secondary and borderline. VO2max was a secondary outcome, multiplicity was adjusted only for the primary outcome, and the trial was…
Siu et al. randomised 315 inactive adults with overweight and central obesity to once-weekly HIIT (H1), thrice-weekly HIIT (H3) or a health-education control (CON), 105 per arm, for 16 weeks, with follow-up at week 32.
Both HIIT arms did 75 min of interval training a week on a treadmill, supervised in groups: a 25-min bout of four 4-min intervals at 85-95% HRpeak with 3-min recoveries at 50-70%. H3 did one bout per session three times a week; H1 did all three bouts in one session with 15-30-min breaks between them. The first 4 weeks built up to the full dose.
The primary outcome, DXA fat mass at week 16, fell in both HIIT arms against control (−0.8 and −1.0 kg) with no difference between them.
VO2max, a secondary outcome, rose against control by 1.6 mL/kg/min with one session and 2.6 with three; the direct contrast, three vs one, was +1.1 mL/kg/min (95% CI −0.0 to 2.1, p = 0.058), and +1.4 (0.3 to 2.4, p = 0.0109) after adjusting for sex, age, BMI, socioeconomic status, attendance, habitual activity and diet.
At week 32, 16 weeks after the programme stopped, VO2max was still higher than control in H3 (+1.4) but not in H1 (+0.9, CI crossing zero), and fat mass no longer differed. Attendance was 86% (H1) and 83% (H3); there were no study-related adverse events.
Our reading: in a Tier 0 population one long interval session a week raised fitness, and three sessions of the same total time raised it a little more and held it longer; for athletes it supports "one good session beats none in a bad week" more than any weekly count.
Protocol
Design: single-centre, assessor-blinded, parallel-group RCT, 1:1:1, block randomisation (block sizes 6-12) by a blinded statistician, sealed opaque envelopes; CONSORT and multi-arm extension; enrolment 1 September 2021 to 23 November 2023, data collection completed 7 September 2024; IRB University of Hong Kong/Hospital Authority Hong Kong West Cluster (UW20-066); ClinicalTrials.gov NCT04887454; protocol published.
Participants: 495 assessed; n = 315 randomised (105 per arm). Eligible: ethnic Chinese adults ≥ 18 y, BMI ≥ 23, waist ≥ 90 cm (men) or ≥ 80 cm (women).Excluded: contraindications to exercise testing, conditions preventing brisk walking, regular exercise training in the past 3 months, weight-loss interventions in the past 6 months; 48 were excluded over concerns about their ability to adhere to thrice-weekly HIIT. 204 women (64.8%); age 48.2 (12.5) y; BMI 28.7 (3.8); waist 94.8 (9.3) cm. Retention 94% at week 16.
HIIT arms: 12 high-intensity intervals a week on a motorised treadmill, supervised by a certified trainer: four 4-min intervals at 85-95% HRpeak with 3-min active recoveries at 50-70% HRpeak = one 25-min bout; 5-min warm-up and cool-down at 50-60% HRpeak. H3: one bout per session, three sessions a week. H1: three bouts in one session with 15-30-min breaks between bouts.Weeks 1-4 familiarisation (50% of the weekly dose by week 2, 100% by week 4) with bodyweight lower-body strengthening and stretching. Intensity from a baseline cardiopulmonary test; Polar A300 with OH1 sensor; speed adjusted first, then incline.
Control: a 150-min group health-education class every 2 weeks, time-matched for attention.
All arms: asked to keep usual activity and diet; accelerometry (7 days) and 3-day weighed food diaries at each assessment. No access to the programme after week 16.
Outcomes: primary — change in DXA total fat mass, baseline to week 16; secondary — fat mass at week 32, fat percentage, lean mass, body weight, BMI, waist, blood pressure, fasting glucose, lipids and VO2max (COSMED Quark CPET) at weeks 16 and 32.
Analysis: intention-to-treat, multiple imputation (25 imputations), generalised estimating equations with baseline as covariate; linear contrasts H1 vs CON, H3 vs CON, H3 vs H1; Holm-Bonferroni only for the primary outcome; sensitivity analyses adding sex, age, BMI, socioeconomic status, attendance, habitual activity and diet. Sample size: 84 per group for 90% power at α = 0.01 to detect d 0.6 for fat mass vs control, plus 20% attrition.
Funding and interests: General Research Fund of the Research Grants Council, Hong Kong University Grants Committee (17105920, 17110722, 17112223) and the Seed Fund for Basic Research of the University of Hong Kong, awarded to P.M. Siu; "The funders had no role in the design, data collection, data analysis, and reporting of this study." The authors declare no competing interests.
Key Results
| Outcome (adjusted mean difference, 95% CI) | H1 vs CON | H3 vs CON | H3 vs H1 |
|---|---|---|---|
| Fat mass, week 16 (primary), kg | −0.8 (−1.4 to −0.2), p = 0.0107 | −1.0 (−1.6 to −0.5), p = 0.0003 | −0.3 (−0.8 to 0.3), p = 0.3586 |
| VO2max, week 16, mL/kg/min | +1.6 (0.5 to 2.6), p = 0.0035 | +2.6 (1.6 to 3.7), p < 0.0001 | +1.1 (−0.0 to 2.1), p = 0.0583 |
| VO2max, week 16, sensitivity model | +1.5 (0.5 to 2.5) | +2.9 (1.9 to 3.9) | +1.4 (0.3 to 2.4), p = 0.0109 |
| VO2max, week 32 (16 weeks after stopping) | +0.9 (−0.3 to 2.1), p = 0.1296 | +1.4 (0.2 to 2.7), p = 0.0219 | +0.5 (−0.7 to 1.8), p = 0.3722 |
| Waist, week 16, cm | −1.9 (−3.1 to −0.8) | −1.6 (−2.8 to −0.4) | +0.3 (−0.8 to 1.5) |
| Fat mass, week 32, kg | −0.3 (−1.0 to 0.5) | −0.2 (−0.9 to 0.5) | 0.0 (−0.7 to 0.8) |
VO2max means (mL/kg/min): CON 26.8 → 26.6, H1 26.4 → 28.0, H3 25.9 → 28.7 at week 16 (about +6% and +11% in the HIIT arms, the authors' figures); week 32: 26.4, 26.9 and 27.3.
Fat mass relative change, week 0-16 (mean, SD): CON +0.2% (7.09), H1 −2.6% (7.24), H3 −3.5% (6.25) — the spread in the HIIT arms was no larger than in the control arm, so the data give no sign of a distinct responder group.
Body weight and BMI: only H3 fell against control at week 16 (weight −0.8 kg, −1.5 to −0.2).
Unchanged: lean mass, systolic and diastolic blood pressure, fasting glucose, total, HDL and LDL cholesterol and triglycerides at both time points per the text; Table 2 does print one glucose contrast below 0.05 (H1 vs CON at week 16, −0.14 mmol/L, p = 0.0454) under a non-significant group effect (p = 0.2124).
Fidelity: mean attendance CON 88%, H1 86%, H3 83%; ≥ 70% attendance in 89%, 86% and 81%; intensity criteria met by 74% (H1) and 79% (H3); duration criteria 98% and 100%; energy expenditure 488 and 480 MET-min/week.
Time cost: weekly commute plus exercise about 4.5 h (H1) vs 8.2 h (H3).
Safety: no study-related adverse events.
Mechanisms
Measured: with weekly interval time matched (75 min, about 480-490 MET-min), three sessions raised VO2max by about 1 mL/kg/min more than one and kept a significant gain over control 16 weeks after stopping; fat mass responded alike.
Authors' interpretation: the once-weekly ("weekend warrior") pattern reduces time barriers, mostly by cutting commuting; more frequent sessions may give more continuous residual effects of the last bout, relevant for glucose control, which was not seen here because glucose was normal at baseline.
Inference: the H1 session holds three interval bouts with 15-30-min rests, so it is closer to a long multi-set workout than to a single 4×4; whether a single 4×4 once a week does the same was not tested here (REF-1288 suggests it moves VO2max little in active young men).
Practical Application
Is one interval session a week enough? In 315 inactive adults with central obesity, 75 min/week of 4×4 intervals as one visit (three 25-min bouts with breaks) or three both raised VO2max vs control; 3 vs 1: +1.1, p = 0.058. A, Tier 0.
A bad week: one long quality session is a real floor, not a wasted week (inference from an A trial in Tier 0 adults who did it every week for 16 weeks): the once-weekly arm did three 4×4 bouts in one visit and still gained fitness and lost fat against control.It sits beside, not inside, SYNTH-75: that guide's beginner default is one short interval set and it lists sets beyond ~23 min as adding little, whereas this once-weekly session held 75 min of intervals; SYNTH-26's "a floor for a bad month, not a program" is said of minimal sprint doses, while here once a week was the whole 16-week programme.
Spread the same work over more days when you can (secondary outcome, borderline): three sessions gave about 1 mL/kg/min more VO2max at week 16 and the only gain still present 16 weeks after stopping. Do not read "once is as good as thrice" from the primary fat-mass result.
Not from this paper; the base's own practice: for a trained endurance athlete the weekly count follows SYNTH-26's decision table (usually two quality sessions on four to seven sessions a week, 2-3 on eight or more; recovery decides the actual count); this trial says nothing about athletes beyond the direction of the frequency effect.
Limitations
The frequency contrast is secondary and borderline. VO2max was a secondary outcome, multiplicity was adjusted only for the primary outcome, and the trial was not powered for H1 vs H3; the three-vs-one VO2max difference was p = 0.058 in the main model and p = 0.0109 in the adjusted one.
Tier 0 population. Inactive middle-aged Chinese adults with central obesity and VO2max about 26 mL/kg/min; the authors limit generalisability to this group.
Selection. 48 people were excluded for doubts about their ability to adhere to three sessions a week, so the trial's adherence reflects people judged able to keep it up.
Unequal session shape. The once-weekly session was three bouts with long breaks, a long visit; the comparison is "same weekly interval time, one visit vs three", not "one 4×4 vs three 4×4".
Diet and activity between assessments not monitored (the authors' limitation), though measured energy intake and expenditure changed little.
Funding and interests: Research Grants Council of Hong Kong and University of Hong Kong seed fund; no competing interests declared.
How this entry was checked and changed
Why this level
large pre-registered RCT, n = 315 randomised and analysed, read from the full text (open access, CC BY-NC-ND). The frequency findings used here (VO2max, three vs one session) are secondary outcomes without multiplicity adjustment, and the trial was not powered to compare the two HIIT arms (the authors say so); the level is the design's, the population is Tier 0
Full text (checked 2026-10-27)
Participants were randomly assigned in a 1:1:1 ratio to once-weekly HIIT (H1), thrice-weekly HIIT (H3), and control (CON) groups (n = 105 per group); the interventions lasted for 16 weeks (full text)
H1 vs. CON fat mass: −0.8 kg [95% CI: −1.4 to −0.2], P = 0.0107; H3 vs. CON: −1.0 kg [95% CI: −1.6 to −0.5], P = 0.0003; H3 vs. H1: −0.3 kg [95% CI: −0.8 to 0.3], P = 0.3586 (full text)
VO2max week 16: H1 vs control 1.6 (0.5 to 2.6), H3 vs control 2.6 (1.6 to 3.7), H3 vs H1 1.1 (−0.0 to 2.1) 0.0583 (full text)
VO2max week 32: H1 vs control 0.9 (−0.3 to 2.1) 0.1296; H3 vs control 1.4 (0.2 to 2.7) 0.0219 (full text)
The H1 group performed the 25-min HIIT bout three times in one session, with breaks lasting 15 to 30 min in between each HIIT bout; four 4-min high-intensity intervals at 85–95% HRpeak (full text)
The mean attendance rates in the CON, H1, and H3 groups were 88%, 86%, and 83%; no study-related adverse events (full text)
The other reasons were concerns about participants’ ability to adhere and comply with thrice-weekly high-intensity interval training (n = 48) (full text)
Supported by the General Research Fund of Research Grants Council (RGC), Hong Kong University Grants Committee (project numbers: 17105920, 17110722, and 17112223), and the Seed Fund for Basic Research of the University of Hong Kong (full text)